Corra v2.0 User s Guide

Transcription

1 Corra v2.0 User s Guide Corra is an open source software Licensed under the Apache License, Version 2.0 and it s source code, demo data and this guide can be downloaded at the This user guide is written by Micheleen Harris (mharris@systemsbiology.org) and Mi-Youn Brusniak (mbrusniak@systemsbiology.org) June 10, 2010 Mi-Youn Brusniak Page 1

2 1. Introduction Corra is a single, user-friendly, informatic framework, that is simple to use and fully customizable, for the enabling of LC-MS-based quantitative proteomic workflows of any size, able to guide the user seamlessly from MS data generation, through data processing, visualization, and statistical analysis steps, to the identification of differentially abundant or expressed candidate features for prioritized targeted identification by subsequent MS/MS. In the first published version of Corra software with the paper was v 1.5 in 2008 and since then, there were more update in the pipeline. In detail, Corra v1.5 pipeline ended by generating target list from statistical analysis. Corra v2.0 added additional feature extracting alignment tools as well as customized target list generation and annotation step using target LS-MS run. This guide uses the yeast gene knock out example used in Corra paper to illustrate the step of using v2.0 extended pipeline steps. 1.1 Login Website: Ask administer in your institution which server the Corra is deployed to and ask Corra admin to add your account. For this guide, we will use guest account. The URL should be something like the following Once logged in, click New to create a new project and give it a name (here it is CourseExample ). Mi-Youn Brusniak Page 2

3 1.3 Choose the instrument type under drop down menu MS Instrument 1.4 Adding Data. Your data must be in mzxml format (if not, there are several converters from RAW data to mzxml, such as ReAdW and mzwiff). Click Add next to mzxml files to add mzxml formatted data to the project (required before you save the project). Select the mzxml files from the drop-down menu with which you want to run Corra (you can hold down the Shift key to select a group of files). Then press Save and reopen Project Setup by pressing Edit. Mi-Youn Brusniak Page 3

4 1.5 Defining Conditions, Sample IDs, Replicates and Time points. Make sure you have clicked Edit to continue setting up the project. Click on Condition_1 or Condition_2 to rename these labels. If you wish to add any more conditions, click Add. Check the files to label and use the dropdown menu to select the condition label appropriate for this group of files. Each group of files (e.g. replicates belonging to a particular biological group) should share the same Sample ID. Assign a numerical ID by clicking on a number in the Sample ID column as shown left side. Define replicates using the drop-down menu in the MS Replicates column. If you have more than 3 replicates increase the replicate count by clicking on the number next to Max. Replicate Count. Mi-Youn Brusniak Page 4

5 If you have more than one defined time point, add it by clicking the Add button next to Time1 and rename by replacing t_1 or t_2 etc. Then specify them in the Time Point column using the dropdown menu. Don t forget to Save your work! Alternatively, you could setup the project by importing a Sample Information File. This is useful if, say, you have a similar project with many mzxml files, as entering all of the setup information by hand could be a rather long process or you can use Copy project option which will create a new project with current project setup page. This Copy option can be used to analyze data using alternative Corra pipeline options. 2. Feature Picking 2.1 Click on Feature Picking, then Edit 2.2 Program for feature picking Select the desired program (e.g. SpecArray is used for TOF-MS data and SuperHirn /msbid for FT-MS/Orbitrap) Mi-Youn Brusniak Page 5

6 For the Feature Picking step it might be useful to view the mzxml file(s) data. Using Pep3D, a.png (image) file can be created and viewed in a generic graphic viewer. For example, using the SuperHirn program, the elution window is set by default to begin at 12 and end at 87 minutes. Viewing the mzxml file in a program like Pep3D can help you decide if you wish to exclude (or include) parts of the experiment based on how the elution profile looks (Pep3D is a viewer of LC-MS or LC-MS/MS data in a general 2D gel-like format). Mi-Youn Brusniak Page 6

7 2.3 Set Parameters or import a parameter file 2.4 Click Run Feature Picking Note: Text in yellow indicates a process that is currently running and text in green is a process which has completed successfully. Text in red indicates an error has occurred and Corra log files may be referenced for further information. When Feature Picking is done, you can scroll down to view the resulting feature counts for each input file. Note: The FT peak detect MS1 intensity min threshold could be increased in the case where there are too many features and/or you desire the subsequent runs to be faster (adjust the parameter and rerun Feature Picking). Here is a picture of the result of the feature counting: Note: These pictures can be downloaded as a.pdf file through the link below this graph. 3. Alignment Mi-Youn Brusniak Page 7

8 3.1 Click on the next step, Alignment. Click Edit to setup the Alignment parameters. 3.2 Select a program (this should correspond to the program selected during Feature Picking) 3.3 Parameters Mi-Youn Brusniak Page 8

9 Adjust parameters to meet the specifications of your analysis and then click Run Alignment. Note: It might be a good idea to start with a value of 5 for the MS1 retention time tolerance. 3.4 Alignment Results An APML (Annotated Putative Peptide Markup Language) file is created and maybe downloaded (by clicking the APML link in red) and viewed in an APML viewer comes with Corra ( Mi-Youn Brusniak Page 9

10 This will help the user to view in a graphical way, the amount of aligned features. See next section for details about the APML Viewer APML Viewer Open an.apml file in the viewer to see the aligned features in a m/z vs. Tr plot. Try this: In the Plotting Tool Bar, go to Selected Plot View -> Times Aligned View and click on a point in the graph to get a dialog box which shows the aligned features for that point (in this case there are three features aligning): 4. Statistical Analysis 4.1 Setup Statistical Analysis Mi-Youn Brusniak Page 10

11 Click on Statistical Analysis and Edit. The program in use is a collection of R modules called CorraStatistics.R. Set the B-Statistics Cutoff (B = -[ log odds ratio]) or use the default of 2.2. Here we change it to 0. Usually, having a N/A Replace Method of none is satisfactory. The N/A Replace Method is to be used if you wish to fill in missing features with a value, either a minimum value or userdefined value. Use the drop-down menu to select the type of N/A Replace Method to use. Select the comparisons to be calculated (red circle on figure below). Save the setup and the Statistics step will begin. 4.2 Results The result of the statistical analysis is displayed in two plots, a volcano plot and hierarchical cluster (unsupervised) as shown below. The red dots are features which are found in some of the samples, but not all and have a Log Odds ratio greater than the value set in the B-Statistics Cutoff field (I set it to 0, here, but the default is 2.2). The blue dots represent features that were found in all of the samples with a Log Odds greater than the default or user defined limit. A Log Fold Change which is negative, indicates that a feature is more abundant in Condition 2 (here, WildType) whereas a positive Log Fold Change indicates that the feature is more abundant in Condition 1 (here, DeletionStrain). Mi-Youn Brusniak Page 11

12 The Volcano plot: The Hierarchical cluster tree: A tab delimited file (.tsv) is created for the aligned features and can be downloaded by clicking Differentially Expressed Feature List (.tsv) link below the Volcano Plot. The data contained in the.tsv file comes from an analysis using CorraStatistics.R (Bioconductor) as a backend. It is used as input in the following step, Inclusion List Builder. The Differentially Expressed Feature List (.tsv) file is shown below, opened in MS Excel. Note in addition to the data there is statistical information such as logfc (log fold change), p-value and B value. 5. Incl usio n List Buil der Mi-Youn Brusniak Page 12

13 Go to the Inclusion List Builder section. (Note: Inclusion List Builder depends upon the Statistical Analysis step which must be completed successfully). Click Create Inclusion List and give it a name. Click F to add a filter and Add Filter. Select a type of filter by clicking on the + sign (circled in blue in next figure) and using the drop down menu. To delete this filter click on the - sign (circled in orange). A #LC aligned filter might be useful if you wish to focus on the number of features aligned across all samples a certain amount of times. A Mean Intensity filter might be useful when features with low intensities wish to be excluded (Note: mean intensity is the log 2 Intensity of a peak). Mi-Youn Brusniak Page 13

14 Hit Save Inclusion List and it will show how many total features and filtered features there are, plus some information such as min and max m/z (filters may be applied to limit these as well). Press F again to close the filter menu. allow the machine to focus on certain parts of the run and not focus on others. Click the red S to modify the segment settings. Segments can be useful when using the Thermofinnagin machine as these can be programmed into a target run. They can Segment Length is the window so to speak (minutes). The segment overlap is how many minutes one wishes to expand the window before and after the segment. First Segment Start is usually just zero, but First Segment End is important as this first segment might capture parts of the run (usually at the beginning) where nothing very informative is happening. The Min. Features per Segment and Max. Features per Segment might be useful to play around with if there are too many features or too few. Click Create Segments. To view the result of segmentation click on View Segment Summary and something like this should be displayed: Mi-Youn Brusniak Page 14

15 In this case I only had 100 features to begin with so there are very few features in my segments so I might try to increase my segment length. You may save list inclusion list by clicking Export to Excalibur. 6. Target Feature Annotation This module is to be used after MS/MS identification of peptide fragments to add sequence (and other protein descriptions) annotations to the original sample spectra, beginning the process of identifying proteins of interest. These could be the focus of future DDA or SRM analyses. Target Feature Annotation annotates the statistical analysis output data (volcano plot data) based on the m/z values in a.pep.xls input file (provided by user). 6.1 Add an input file somefile.pep.xls Mi-Youn Brusniak Page 15

16 Note: This input file can be created using a pepxml viewer to convert a.pep.xml file to.pep.xls (e.g. PepXML Viewer part of the TPP, see tutorial at This is a screenshot from PepXMLViewer (uses PeptideProphet analysis): The input interact.pep.xls file should have at least these headers (but you may add more, like the index and spectrum for instance): assumed_charge MZratio peptide retention_time_sec protein In order to run Target Feature Annotation you must Add an.xls (e.g., interact.pep.xls) file which has all of the possible annotations that may be queried and added to your data (volcano.tsv file actually). See next figure for adding a xls file. Mi-Youn Brusniak Page 16

17 You may adjust the m/z tolerance in ppm which is set at a default to 25 ppm. Also, you may wish to adjust the rt tolerance in minutes. These are worth playing with if you do not get very many features annotated. Then, hit Run Target Features Annotation. 6.2 Results of Target Feature Annotation Once Target Feature Annotation has run, you will have an annotated volcano.tsv file from section 4. Mi-Youn Brusniak Page 17

18 At this point, you may click Download Target Feature Annotation TSV, (circled in red) a file which looks similar to the Differentially Expressed Feature List (.tsv) from section 4, but with additional information including peptide descriptions for some of the features. Below, the annotated.tsv file is shown opened in MS Excel. You may also download the IPI file TSV which just contains just the features which are associated with IPI(s). Mi-Youn Brusniak Page 18

19 6.3 PIPE2 You may load the results of Targeted Feature Annotation into PIPE2 by clicking Load Proteins to Pipe2. Alternatively, you can copy and paste your IPIs, ORFs or other feature identifiers into PIPE2 to map them to several other databases, providing additional information about these important features. The PIPE2 link is here: Note: you must have firegoose extension installed in your computer when using Mozilla Firefox browser ( Here is a screenshot of our yeast DeletionStrain and Wildtype data into PIPE2 after Target Feature Analysis (remember we found 52 aligned and annotated features). This links to a tutorial of PIPE2: Mi-Youn Brusniak Page 19

20 7. Trouble Shooting Problem The mzxml file format is invalid PIPE2 button does nothing in firefox browser Fix try using indexmzxml.exe to correct errors Install firegoose extension Mi-Youn Brusniak Page 20